Generalized Linear Varying Coefficient Model with Data Missing at Random

• Activity Number: 353
• Type: Contributed
• Date/Time: Tuesday, July 31, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Section on Nonparametric Statistics
• Author(s): Jianwei Chen*+ and Qian Xu
• Companies: San Diego State University and San Diego State University
• Keywords: Generalized linear varying-coefficient models ; quasi-likelihood imputation estimator ; missing data ; the mean function.

Abstract

The generalized linear varying-coefficient model is an important extension of the generalized linear model. The model structure allows the coefficient to be a curve function with different time. Since local quasi-likelihood estimation is useful for nonparametric modeling in generalized linear models, we extend three estimation methods from it for the generalized varying-coefficient model when there are data missing at random: the local quasi-likelihood estimator using only complete-data, the local weighted quasi-likelihood estimator and the local quasi-likelihood estimator with imputed values. We develop the local quasi-likelihood imputation methods for estimating the mean function of the response variable. Our simulation results show that the proposed imputation methods perform better than both the one based on complete-case data only and the weighted method.